The main aims of this study are to investigate the relation of diet to ovarian cancer, and to investigate whether the age at use and 'dose' of oral contraceptives (OCs) affects their protective effect against ovarian cancer. These aims will be accomplished by a population-based case-control study. We will be using a questionnaire-based approach with 750 case- control pairs, with a laboratory component (measuring galactose metabolism) for 50% of the subjects (see below). In-person structured interviews will be conducted using a month by month calendar for reproductive events with a photograph album of OCs and other aids to facilitate recall. All cases will have their diagnostic slides blindly reviewed by a single pathologist to attain uniformity of diagnosis. Validation will be undertaken of OC use and other selected aspects of the subjects' medical histories. The main dietary issues are whether animal (or saturated) fat consumption and/or lactose consumption are positively related to ovarian cancer risk. These questions will be addressed by a validated dietary food frequency questionnaire. The lactose issue will be further addressed by investigating the relationship between lactase persistence and risk, and by studying galactose metabolism. Lactase deficiency will be investigated on all subjects through a questionnaire approach, and the relationship of the questions asked to actual lactase deficiency will be investigated in a sample of controls. Serum galactose acting directly on the ovary has been proposed as the mechanism of the proposed relationship of lactose consumption to ovarian cancer risk; this is the reason for studying galactose metabolism. A recent overview of a number of epidemiological studies of ovarian cancer provided quite strong evidence that age at menopause is effectively unrelated to ovarian cancer risk. This, and other observations suggesting that the known protective effects of OCs against ovarian cancer may be related to OC dose and age at use, provides real discrimination between various hypotheses of ovarian cancer 'etiology'. We will investigate this further by collecting detailed data on OC brand and age at use, and relating the protection found to the protective effects of pregnancies at various ages. An in-depth understanding of the protection afforded by OCs is important for many reasons, in particular it is vitally important to know whether modem low-dose OCs are continuing to provide significant protection against this important cancer of women.